An Integrated Framework for Crew - Centric Flight Operations

Abstract

This paper presents an integrated framework for crew-centric flight-deck operations within the FP7-EU funded ACROSS project. ACROSS is developing, integrating and testing new solutions to reduce pilots’ peak workload and stress, supporting them when dealing with difficult situations, thus enhancing safety and performance. ACROSS presented a number of human factors challenges: (1) diverse technologies being designed simultaneously and in parallel, (2) multiple partners throughout Europe with different needs and design philosophies (3) production of multiple technologies relating to different parts of the flight operations process. The global human factors challenge was to produce an integrated human factors approach that would facilitate the best outcome for ACROSS. To meet this challenge a crew-centric framework for flight operations was designed. This paper presents the framework itself, the development process and an illustration of the concepts behind it.

1 Introduction

This paper presents an integrated framework for crew-centric flight deck operations developed within the FP7 EU-funded ACROSS (Advanced Cockpit for the Reduction Of Stress and Workload) project. ACROSS is developing, integrating and testing new flight deck solutions to reduce pilots’ peak workload and stress, supporting them when dealing with difficult situations, thus enhancing safety and performance. These solutions will facilitate the management of the peak workload situations that can occur during a flight and allow a reduced crew to operate within a limited number of well-defined conditions. ACROSS has three main objectives:

• 1. New Cockpit Solutions for Peak Workload Situations. The ACROSS project will contribute to a cockpit environment that mitigates the impact of crew workload peaks in the flight deck and ensures that pilots have the opportunity to address all relevant issues in a timely and effective manner;

• 2. New Cockpit Solutions for Reduced–Crew Operations. ACROSS will develop and integrate cockpit-based technologies that allow the remaining pilot to safely manage the flight.

• 3. Identifying Open Issues for Possible Single Pilot Operations. ACROSS will identify aspects that currently prevent the reduction of the crew to a single pilot.

Single-pilot operations in all conditions are considered a long-term evolution that is not in the scope of ACROSS research & technology developments. The ACROSS consortium considers single-pilot operations as a case study that stimulates innovation and facilitates the identification of solutions that could be used to improve the current safety level in situations of peak workload and reduced crew. Conversely, any solutions developed to manage peak workload and reduced crew situations may be considered for possible single pilot operations in the future.

1.1 Integration as a Challenge in ACROSS

The technology development work in ACROSS is organized around six main pillars (1) Aviate, (2) Navigate, (3) Communicate, (4) Manage Systems, (5) Crew Monitoring and (6) Crew incapacitation. Technologies are developed to support crew in performing their operational duties along the functions of these pillars. From a human factors perspective, the challenge was to ensure that all partners had the same understanding of the human factors issues and were designing their technologies and evaluating the concepts in a consistent way throughout the project. If partners had differing understanding of concepts such as workload, situational awareness, automation, incapacitation, it would be unlikely that the technological solutions would seamlessly support the adaptive needs and complex relationship between human and automation on the flight deck. The integrated functional concept outlined in this paper was informed by the needs highlighted during facilitated interactions between pilots, human factors experts and the technology developers in the project consortium. Five particular concepts were identified by human factors experts as outputs of a human factors integration workshop: (1) Crew as manager of operation (2) Crew Incapacitation (3) Progressive automation (4) Emergent automation issues (5) Distributed team working. These concepts were seen as drivers for what a crew-centric framework for flight operations should support.

1.2 The Crew as Manager of Operation

Flight-crew being at the center of operations was considered a fundamental part of the ACROSS philosophy. The multi-disciplinary working groups organized by the human factors team all stressed that the pilots must be at the center of operations and must have the final authority about decisions made on the flight-deck. For safe and effective decision-making to take place on the flight deck, it is imperative that pilots have the authority for making final decisions (Harris 2007) and can choose not to take the automated support options if necessary. Parasuraman and Wickens (2008) highlight the importance of clear understanding of whether systems should be user controlled or system driven.

1.3 Incapacitation

The extent to which cockpit crew, both individually and collectively, have the capacity to meet the demands of flight operational tasks is a key factor in flight safety. Pilots are said to be incapacitated when their ability to control the aircraft effectively and perform flight operations is affected adversely by physiological and/or psychological factors (Taneja and Wiegmann 2002). Crew incapacitation is not a very rare phenomenon. For instance, Evans and Radcliffe (2012) reported that the annual incapacitation rate for UK pilots for 2004 was 0.25 %, (approximately 1 in 400). Another study involving pilots from a British airline showed that 45 % of respondents admitted that they were suffering from significant fatigue and one in five indicated that they abilities were compromised in flight more than once per week (BALPA 2011). Clearly, crew incapacitation affects flight safety and thus it is a major concern in the aviation sector generally. Due to its focus on workload detection and subsequent management, ACROSS has one technology development pillar which is devoted to detecting incapacitation and mitigating its effects (Liston and McDonald 2014). Generally speaking, various forms and degrees of incapacitation can be monitored using physiological, behavioural and self-report measures. Subsequently, the data can be used to quantify the impact of incapacitation on the flight operational process and the mission outcome. Of greatest concern however, is full crew incapacitation. Whereas this phenomenon is rare the impact of such incidents is generally catastrophic. Moreover, in situations where crew numbers are already reduced, the number of incidents involving a fully incapacitated crew will likely increase significantly. Thus, a variety of means are being explored in ACROSS for controlling the aircraft either automatically or remotely. For these to work effectively, reliable and valid data needs to be collected and analysed so that any decision to declare a crew fully incapacitated is justified.

1.4 Progressive Automation

Sarter et al. (1997) outline some unintended consequences and pitfalls of increased automation, such as assuming that increasing the automation on the flight-deck will automatically simplify the task, free-up precious human resources reduce ambiguity and human error when in reality the opposite can occur. Simply adding more technology does not necessarily mean that the human is better catered for in terms of effective support on the flight-deck (Dekker 2004). Stein et al. (2009) describe progressive automation as “an approach that has a better philosophical base for what automation seeks to achieve and a more human-centred approach to avoid the most adverse human-factors consequences of automated systems and provide a better planned, progressive introduction of automated aids in step with user needs”. (Chap. 21, p. 29).

1.5 Emergent Automation Issues

Parasuraman and Wickens (2008) define adaptive automation as “An alternative approach that attempts to reap the benefits of automation while minimizing its costs is to vary function allocation during system operations”. (p. 517). To date, true adaptive automation has largely been evident in the military domain (Harris 2011) and there is little comparable technology available today in commercial aviation. Mosier and Fischer (2010) outline a comprehensive overview to appropriate decision-making support to ensure the correct balance between human and automation. They stress that:

“Decision aids need to be adaptive not only in the when of aiding, but also in the how”. (p. 239).

Parasuraman and Wickens (2008) outline the benefits of both adaptive and adaptable automation. They stress that one solution to the questions of “when” and “how” that Mosier and Fischer (2010) pose would be to adopt a common language that can be understood by both the human element and the automation so that tasks can be delegated as necessary. Mosier and Fischer also highlight emergent issues such as attention guidance, transparency, the level of intervention, facilitation of appropriate cognition, teamwork and shared mental models are considered. Also, decision support in multiple modes (such as visual, auditory, haptic feedback) of decision support should be considered for future research.

1.6 Distributed Team Working

Malakis and Kontogiannis (2014) outline the challenges surrounding sensemaking when the flight deck and air traffic controllers have similar but not identical information available. This would be particularly relevant for the development of ground systems designed to support flight crew in the case of incapacitation. However, Salmon et al. (2008) highlight that individual team members can have different situational awareness of the same situation and still perform successfully as a team. Sorensen and Stanton (2015) demonstrate the importance of schemata in ensuring compatible situational awareness amongst team members. It was considered fundamental to the multidisciplinary team that provided operational feedback in ACROSS to have technology and interfaces that would support the optimum individual and compatible team situational awareness.

2 Methodology - the Development of the Crew-Centric Framework for Flight Operations

2.1 Human Factors Group Integration Workshop

At the end of the first year of the project, it became apparent that human factors challenges in ACROSS needed to be addressed in a more global and systematic manner. An integration workshop was held and was attended by the wider HF group in ACROSS. This is composed of representatives from industrial partner organizations, technology developers and research organizations. A number of these representatives are also pilots who make up the internal pilot group in the project. The workshop contained an exercise to consider HF as an integrated conceptual framework with crew at the centre of operations. The group were asked to consider what kind of conceptual framework could be built for an integrated HF approach in ACROSS? They were given the brief that crew were at the centre of flight operations, i.e. they were managers of the flight operation. For a number of scenarios, this was broken down into: What would crew need? When would they need it? How would they get the information they needed “to manage the operation”. This exercise produced 5 main drivers which would support an integrated crew-centric framework: (1) Crew as manager of operation (2) Crew Incapacitation (3) Progressive automation (4) Emergent automation issues (5) Distributed team working. These drivers were used to design the framework which can be viewed as the 1st stage on the journey towards achieving a global and integrated HF approach in ACROSS.

2.2 Integrated Human Factors Teams – Internal and External Human Factors Working Groups

The HF integration workshop was also instrumental in setting up of a formal network of HF representation within each partner organization. Human factors constitute a major driving force in ACROSS. ACROSS has 35 partners based throughout Europe and the many technologies produced are diverse in nature. The Integrated Human Factors Teams (IHFT) were set up in order to ensure integration between partner organizations and their work and consistency in their approach to human factors issues. A separate IHFT was set up for each of the six main pillars in ACROSS (1) aviate, (2) navigate, (3) communicate, (4) manage systems, (5) crew monitoring, (6) crew incapacitation. The IHFT were thus constructed as a solution to the problem of matching the demand for HF support with the supply of HF expertise in the project. Each IHFT contains two sub-groups, an internal working group consisting of technology developers and “in-house” HF experts from industrial partners and an external working group consisting of HF experts from within the project. The IHFT meet on a regular basis to discuss progress on design of the ACROSS solutions, integration with other technologies from the other pillars in ACROSS, human factors issues and evaluation of the work to date.

2.3 Integration Activities

The integration activities described herein, represent the groundwork for the second stage in this journey towards being able to evaluate concepts such as workload and situational awareness at a global level. Stage 2 will continue to the end of the project. The human factors group in ACROSS devised a number of steps to ensure that all partners were working within a common operational picture, that the human was placed at the centre of the design and evaluation work, that there was a common understanding of the challenges faced by the human in the system and that the human factors approach in ACROSS was both consistent and integrated throughout the project.

2.3.1 Conceptual Working Group (CWG)

A working group, consisting of human factors experts from within the ACROSS consortium was formed in order to review emergent conceptual and methodological issues. The CWG served to support conceptual and methodological integration in the project, thereby ensuring levels of consistency and quality that are necessary for effective validation. The objectives of this group were to: (1) Review project deliverables to check consistency in definition for concepts such as workload, stress, situational awareness, automation etc.; (2) Support technological development partners in identifying how to measure those concepts in a robust manner; (3) Identify human factors experts who are willing and able to either administer tests and/or provide support for other individuals to administer them.

2.3.2 Production of a Human Factors Handbook

An ACROSS human factors handbook was produced to provide a standard reference for common human factors terms, practices, concepts, methods and instruments to be used in the project in order to make this information more accessible to partners in the consortium. The handbook set out the theoretical bases for the human factors approach and also provided practical support to the technology development pillars by detailing the human factors expertise in the project and also outlining a range of HF principles, guidelines etc. which will guide the development of technological solutions. It gathers the more practical human factors elements into a single document, which is intended to be used as an initial point of reference. This handbook fulfills 5 main functions: (1) Provides operational definitions for key concepts; (2) Describes the structure and functions of the ACROSS Integrated Human Factors Teams; (3) Provides an overview of measures and methods that can be used to test key ACROSS concepts; (4) Provides a summary of the automation issues and the automation philosophy and related guidelines; (5) Summarizes the HMI design principles and HMI philosophy in ACROSS. A series of workshops, hosted by the human factors experts in the IHFTs, were conducted in order to demonstrate how to use the Handbook.

2.3.3 Operational Concept Reviews and Evaluation

This integrated concept was presented for review as an operational concept within ACROSS according to the following process: (1) At a specific human factors workshop attended by all members of the wider human factors community in ACROSS; (2) Subsequent periodic reviews and walkthroughs with human factors experts were been conducted via teleconference; (3) A further presentation and walk-through of the concept was carried out at ACROSS convergence meeting with internal ACROSS pilots and technology development managers from the six ACROSS pillars; (4) The concept underwent a formal review process by the ACROSS Project Management Committee and European Commission project reviewers; (5) A final evaluation of this concept will be carried out in relation to all of the six ACROSS pillars.

2.4 Supplementary Activities

Other supporting activities to test the structure of the framework and to ensure that it made logical sense – within the context of design of the individual technologies being developed and the operational reference scenarios used in the project. Process modelling workshops were carried out with technology developers in order to establish how their technologies would contribute to the process model of the flight operations process in ACROSS. A generic process model for a standard flight operations process in commercial aviation. This was updated to include data relevant to each technology pillar. Human factors experts carried these workshops out via teleconference showing the process model on screen and asking for input on human factors issues such as decision-making, loss of service (full loss, partial loss, loss of redundancy, interference) and the impact that this would have a different decision points throughout the mission. These workshops were scenario based and made the use of storyboards to provide structure and time-sequence to the workshop.

In ACROSS, storyboards were used to put flesh on the bones of the reference scenarios and to examine flight operations from a decision making perspective. Such detail is essential in the development of operationally valid process models and also to guide the development of the operational concepts within each technology development pillar. Although the technology development pillars are different in nature, the storyboard structure will provide a consistent structure for describing the narratives, the level of information required, different technologies, conditions and actions. The storyboards remain as living documents which will be updated by the IHFTs. The storyboards query flight operational scenarios from different points of view (different stakeholder views, deviations from standard operating procedures, non-normal operations etc.).

3 The Framework

3.1 The Integrated Framework for Crew-Centric Flight Operations

Figure 1 details the layout of the concept. Each of the ACROSS technologies developed in each of the main ACROSS pillars have an input to the application layer at the bottom of the figure (i.e. the “novel technology” solutions). Information is organized into functional groupings: Manage Mission; Manage Systems; Manage Crew. This is the integration layer of the diagram. Manage Crew represents the high level organization of operational demand and dependencies for this flight, together with the assessment of crew capacity and the management of these dimensions throughout the operation.. The integration layer represents the integrated functions that support the crew in managing the aircraft systems, the mission itself (from gate to gate) and the operation which involves the wider dependencies within the flight operations system at a global level.

Fig. 1.

The integrated functional concept for crew-centric flight operations

The abstraction layer contains the schematic organization of this information into crew-centered dynamic diagrams and other representations. Combinations of these abstractions support the crew in planning, maintaining their shared situation awareness bubble, in concurrent task management, and in briefing, reviewing and reporting. For example, in Manage Mission, the abstraction layer provides a schematisation of the flight along a series of critical points. The information displayed here could be based on waypoint information, time, distance, weather, terrain, fuel consumption etc.

3.2 Illustration of the Integrated Crew-Centric Concept for Flight Operations

The following scenario is used to demonstrate how this integrated crew-centric concept for flight operations can be used to support the achievement of the ACROSS objectives. This scenario was adapted from the set of operational scenarios selected by the ACROSS consortium.

Scenario: A business aircraft crew flew from Europe to the US. The crew flew a local flight (of less than 3 h) within the US with a change of time zone. The crew is now flying back to Europe and is on its third flight shift with respect to Flight Time Limitations. The crew has experienced poor quality sleep due to various time zone changes. In addition to this, the destination airport is not a familiar one and it is raining. ATC puts the aircraft on hold, and then asks if the crew can accept a change of runway for landing. The proposed new runway would allow landing sooner than if they wait for the planned runway, but it is shorter than the planned runway. The crew has to make a decision on accepting or rejecting the new runway proposal from ATC.

This framework for crew-centric flight operations would support the crew in locating themselves within their “situational awareness bubble”. The situational awareness bubble could be defined as the crew’s sense of awareness of where they are in a particular point in the flight operation, their awareness of the status of the aircraft, of their physical surroundings, of what has just happened in the immediate past and their awareness (or anticipation) of what is likely to happen in the immediate future based on all of this information. The framework would further prompt the crew with assistance in the form of a list of prioritized actions and decision support (Fig. 2). The abstraction layer shows that the crew are in the planning phase as highlighted on the left hand side of the figure. They are also presented with a high level analysis of the dependencies, and mitigating factors linked to their operational outcomes.

Fig. 2.

An illustration of a decision support interface

The crew must have a full operational picture to support the situational awareness bubble (individual and/or shared/distributed) for them to be able to make an informed decision. Providing crew with this rich operational picture which includes operational dependencies relevant to flight phase, risk information likely consequences of potential actions is the role of this integrated concept. The dependencies are examined for each node (i.e. task, sub-task) of a flight operations process model. The associated risks for each node are also detailed within the model and can be assessed in order to provide a prioritised set of recommendations and actions that can be carried out to mitigate for the operational conditions that crew are experiencing on the flight deck. For this scenario, crew must be aware of the operational dependencies that could have an impact on a successful landing. These are: (1) Fatigue and sleep loss (2) Unfamiliarity with airport (3) Need to do pre-landing check and briefing (4) Rain (5) Anxiousness to complete flight. These dependencies place constraints on the cognitive capacity and resources that crew have available. Being aware of these dependencies alone is relatively meaningless unless it is contextualised in relation to the potential operational outcomes (i.e. consequences) and the potential actions that could be taken in order to mitigate for the operational constraints (Fig. 2). This integrated concept for crew-centric flight operations will provide a more powerful format of decision-making support with associated risks and consequences (i.e. operational outcomes). This is an improvement on the state of the art currently provided by existing re-routing functions which don’t offer prioritized recommendations with associated operational outcomes/consequences.

4 Discussion

4.1 Integration as a Solution in ACROSS

This framework implemented an integrated HF approach for system development within the ACROSS project. The integration activities also provided consistency of definition and measurement and promoted a shared common perspective putting human at core of design and evaluation. This framework was not only a solution to integration in ACROSS, but also provided means of examining what crew are doing throughout the flight operational process from technological, cognitive and operational perspectives. In the illustrated scenario, the potential actions that crew could take (in this case the decision to land at new runway or original runway) can be prioritized to further demonstrate support for crew in making their decision. For this type of scenario, the action list is likely to be based on risk information and thus safety, but for other nominal scenarios (where safety is has the same low risk factor), this action list could be based upon fuel consumption, distance etc. The crew have the final authority as to which action/decision to select. We have already noted the dependencies above (fatigue and sleep loss, unfamiliarity with airport, the need to do pre-landing check and briefing, rain, anxiousness to complete flight.). Two operational actions would be presented (1- land at new runway, 2- land at original runway as planned). Potential consequences are also presented (e.g. 1- runway overrun, have to land, short excessive breaking; 2- increasing fatigue, crew not fit to make a good landing). The actions and consequences would be colour coded. The green could demonstrate a low-risk factor associated with the condition and the red a high-risk factor associated with the antecedent condition. The risk factors would be based upon previous operational data gathered (FDM data, incident reports etc.) on this particular flight phase and node of the flight operational model. Risk Factors in this scenario were: (1) The crew are unable to make an informed and solid decision, loses time and safety margins; (2) The crew elaborates a poor plan of action, because of fatigue, forgetting or poorly executing one of the required tasks (for example performance recalculation with bad parameters), poorly planning the missed approach conditions etc.; (3) The crew focuses on one aspect only of the situation because of time pressure and fatigue and skips an important piece of information (e.g. runway is contaminated). Crew actions would be fed back into the risk information loop and the system updated for future decision support and planning support for the flight deck. The ground station and concept of a remote pilot (although not fully implemented within the ACROSS project) could also be represented in Fig. 1. The ground station could have an identical interface so that there is heightened distributed situational awareness. This would increase the likelihood of a compatible situational awareness, a good common operational picture and both the ground and remote pilots being brought up to speed as quickly as possible as soon as they are alerted to partial or full incapacitation on board the aircraft.

The design of this framework is important for the purposes of integration, but also from the perspective of providing additional information potential consequences of crew actions. The antecedent conditions are linked to the specific point within the flight operation and the associated mitigating actions. It is of paramount importance to have this information in order to do the following: (1) Further locate this instance in the flight operation to support the crew’s context-rich situational awareness bubble; (2) Allow each instance within the flight operation to be linked to specific dependencies for that instance and subsequent ones within the model; (3) Close gaps within the operational process. The local operational concepts within each technology development pillar do not provide this information as they do not examine the full context of the operation in terms of risk, dependencies, decision points and relevant operational mitigating actions; (4) Allow for risk information to be located precisely within the flight operations process and for the risk information to be linked consequences of potential actions; (5) Allow for risk information to be fed back into the operational model (i.e. updated for the specific point in time of the flight operations. This crew-centric framework could be helpful for current flight deck operations to support crew in managing their own situational awareness bubble and in “managing the operation”. Being able to manage the operation will become ever more important if there is occasion for flight-crew to be operating on their own (through rest-break or incapacitation). Facilitating crew in being able to answer “what if” questions earlier in the process supported by prioritized actions and risk information is key. It is also likely to result in smaller incremental corrective actions as a result.

5 Conclusion

This paper presented methods used to develop an integrated HF approach and demonstrated how this concept relates to the flight operations process. Harris (2011) stresses that human factors in aviation must follow and approach that is “integrated, through-life and systemic” (pp. 7, 321) if it is to succeed in supporting operations from safety, performance and cost perspectives now and in the future. The crew-centric framework as presented here can be considered in-keeping with this philosophy and will be further extended and validated within the lifetime of ACROSS. Stage 2 in this process will develop means for evaluating crew workload in a global and systemic manner. It will also form the basis for future research activities in this area.